DAC Quality: Myths vs Reality

DACs and Digital
1

So, I always thought that the difference with DAC’s was more signifigant than what my real world experience has shown. Between R2R and Delta/Sigma DACs, I’ve always favored the sound of the Delta/Sigma. There justseemed to be more resolution with the Delta/Sigma setup.

Having said that, I’ve also noticed that the differences between DACs seem to be very little. I had bought into the FPGA hype, so I’ve been using Chord/PS Audio DACs. Then, I recently auditioned a Topping DX9, and I was quite impressed with the performance. As a stand alone DAC, connected to a high quality headphone amp, the performance is excellent. I find the sound from the DX9 every bit as good as the high end FPGA DACs, and seems even quieter, with no noticeable distortion.

Found this video about DAC’s on YouTube:

Have to admit, he may have a point.

3 Likes
2

The more experienced people tend to spend the most effort/money optimizing the speaker/room and headphone/ears setup. The differences gained here usually FAR eclipses differences gained by spending same effort in.. say DAC’s.

Even this cheap $240 XiangSheng DAC has no problem being the source of great audio systems.

0326252100a by drjlo2, on Flickr

2 Likes
3

Agreed. I’ve been using the DAC’s in question strictly for headphone listening. My speaker setup is a Devialet Expert Pro, which has an excellent DAC setup.

1 Like
4

There are a few companies that don’t use the major DAC chips in some products. The $300 Schiit Modi Multibit, for example uses an entirely different non-audio DAC chip and designs around that (not a Delta-Sigma). I do think it has a somewhat different sound that does have some family resemblance to the Bifrost 2/64 that I also have, also not an outrageous amount of dollars (or greenbacks, iron men, Simoleans, smackers, or beans but not “quid”).

I don’t think that Schiit is using a similar chip designed for a different (in this case medical) purpose as the video infers. There are delta-sigma DACs that are designed for non-audio use. Read their blurb - makes more sense than I can relate here. And the proof is how it sounds - for that three hundred clams.

Anyway, clearly not doing the 5K Quid thing.

1 Like
5

Hmm. So there’s a lot here that’s just not quite right, even if the sentiment isn’t unreasonable. I think after watching this one might assume the DAC chip alone is responsible for good or bad, and that’s not true. And, I think there are some misconceptions here about what would constitute ‘good’ for people, or like… good to the degree where nobody has a reason to care about the difference, as opposed to answering the question of “can anyone actually hear a difference?”.

Even within the category of DACs and amps that had been deemed [sufficiently transparent] along traditional analyses, that doesn’t mean they actually are in practice. But that’s typically just down to incomplete information.

Broadly speaking though, my perspective on this stuff is very similar to what this person has put forward, and I tend to align with that sentiment. I said this in our live stream a few weeks ago, but the tendency for the audio world to prop up the notion that DACs and amps have a sound, is itself a scam, whether it’s done so knowingly or not. I know not everyone agrees with me on this - even within our team we have some disagreement on this stuff. But after doing the blind ABX testing, I personally have very little reason to care about ‘the sound’ of a DAC or an amp.

That doesn’t mean people aren’t having experiences that lead them to believe these devices sound a certain way, and there are cases where people will genuinely be hearing differences that exist acoustically. But I have yet to come across a scenario where you couldn’t explain those differences in terms of the things we can already measure, or things like impedance relationships with the rest of the equipment and so on.

6 Likes
6

I’m not sure I understand the notion that amps and DACs don’t “have a sound” if they sound different. Certainly, some amps like tube amps very obviously “have a sound”. Other amps are voiced to sound a certain way, for example Nelson Pass is known for designs that intentionally introduce second order harmonics to imitate the sound of tubes. But I mean, I was making a solid state DIY kit and I changed some capacitors and the sound changed, so the idea that amp design doesn’t affect the sound, that components don’t matter, seems to me self-evidently and demonstrably false.

7 Likes
7

I think some people are trapped in DSP and other processing homogeneity that indeed castrates the unique sounds of other devices. But, I’m not going to debate it.

4 Likes
8

Being a typical DAC more than just a DAC chip, does the whole unit make for that difference?

As for measures, I do agree, they matter, and that is really shown in Cameron’s analysis of different DACs and amps. Different units are different. They measure differently.

You can subjectively analise a headphone, and amp, a DAC, but you can really objectively compare each one to similar categorized units.

The interesting debate would be if the subtle differences - I think everybody agrees that they are not night and day - really warrant the price discrepancy :slight_smile:

9

The problem with value based arguments is it’s not absolute, it depends on the person making the purchase and their financial position, and how much they value those differences.

Every difference in this hobby is overstated, it’s the nature of wasting time on forums over analyzing it.

DAC’s in particular are difficult to distinguish between in fast A/B switching tests, I’d argue that’s not really all that interesting, I’ve had multiple higher end DAC’s here at that same time with the ability to keep any or all of them, and with extended listening I have always had clear preferences, for one or another.

Part of the issue I think people who focus on measurements have with subjective assessments, is the language used doesn’t map to measurements. You’ll see midcentic, on something that measures with a flat FR, or terms like faster where there isn’t a measurable difference in slew rate. It’s just difficult to map the language to metrics, that doesn’t discount the perception, it just implies the language is imprecise.

I’m a bit flumoxed that anyone would think all amps sound the same, I’d personally have a hard time telling a difference between the low Sinad, high feedback designs, but outside of that there are very obvious differences to me.

I’ve been designing and building amps for myself for a while now, I do measure everything I can, but short of knowing what bad looks like I’m not sure I could point to the measurements I take and say that’s why it exhibits good bass grip or air or whatever. My current best indicators for a “good sounding” amp are the shape of the harmonic distortion, and how that changes with input level, and the phase response. But I exclusively use low feedback designs with only local feedback (if any) and as a result I’m usually only aiming for 2nd Harmonic distortion in the -70dB range at listening levels.

I often a prototype on strip board with whatever resistor and capacitors I have in my parts bins to validate before ordering PCB’s and higher quality parts, and even with identical circuits and very close measurements, the final designs often sound somewhat different to the prototypes.
But one of the biggest surprises to me was around the impact of physical packaging and internal layout.

5 Likes
10

Right, so the issue is less to do with being able to differentiate two products because one has audible distortion products and the other doesn’t. It’s more how this gets twisted into knots to make people feel one is of a higher quality than another because it “has such a unique sound to it”.

Usually… the idea that a given DAC or amp “has such a special sound to it”, is more to do with things external to the equipment itself. But even in cases where there is genuinely an audible difference from the DAC or amp that holds up in blind tests, that’s usually in the form of spurious or undesirable products being introduced into your signal. So to then celebrate this as “it’s unique sound”, whether in marketing copy or otherwise, is effectively sanitizing what is most likely a worse outcome.

The big exception here is tube amps, and I’ll be the first to admit, this is where the above perspective doesn’t apply. But I also think if you’re getting into tube amps you’re probably already comfortable with audible distortion products in your chain, or at least you probably should be. And I really mean that for myself too… one of my favorite amps, that I call a cheat code for headphones, is the Feliks 25th Anniversary Edition. That creates such a cool effect on most headphones. But better to go in with eyes open on this stuff.

5 Likes
11

IMG_0968

3 Likes
12

I love this guy’s vids. He always seems on the verge of exasperation, which he gets away with because of his accent. :grin: And yes, I strongly suspect he’s right – the other parts aren’t irrelevant but the chip is doing the job and chips aren’t expensive, so a DAC shouldn’t be either.

13

And that’s the crux of his arguement. I admit I was taken in to some extent about the DAC hype, but the Topping DX9 pretty much blew up my past preconceived notions. The unit as a stand alone DAC is excellent. Easily holds its own against more expensive DACs. Since I use the DAC strictly for headphone use, it’s easier to discern what’s going on when comparing. I think the arguement about comparing DAC’s when level matched being very hard to discern is valid.
Amps are a different story. There are indeed sonic differences between amps, but it could be argued that between good quality SS amps level matched, discerning a difference could be challenging. Any differences heard could be attributed to the the amp being stressed to it’s limits with a given specific load. Tube amps will sound different for obvious reasons.
The basic premise here is that there is that modern DAC’s shoud and do sound more alike than different. The need to spend more $$$ on a DAC is questionable at best.

1 Like
14

Thanks for posting this, SSN757.

It would be interesting to hear @GoldenSound’s take on this, since he has much more experience with this type of gear.

Alain often has interesting takes. This is pretty middle of the road for him though, and most audio debunkers.

His main point is a good one: you don’t need to spend alot to get a perfectly transparent sound from a DAC. And I’d tend to agree with that, based on my somewhat limited knowledge of this subject. The question is… is that what every audio enthusiast wants? And I suspect the answer to that is no.

Also, even DACs or DAC chips that are designed to produce a completely transparent sound may be configured to output something different. So it is not necessarily just about the chip or chips being used.

Alain has a followup to this video btw.

He also posted a white paper with the results of his DAC poll here.

I’m not overly impressed with the quality of the recording he chose for this test/study btw. And perhaps that had some influence on his poll results. It might be interesting to see this repeated with some recordings that sound better to begin with.

It would also be interesting to see it repeated with a more random group of participants than just Alain’s own viewers.

1 Like
15

Not sure I agree with everything in this, but a recent video by Super* Review on what’s most/least important in an audio chain (primarily for headphones) also seems to tie in pretty well with this discussion.

1 Like
16

Yeah I’d parse some of them in a slightly different way, but for the most part I agree with Mark Ryan’s ranking here. Certainly I can see the argument for it.

1 Like
17

Heck, I didn’t even use a DAC until the mid 1980s. Didn’t miss it at all.

18

I am clearly behind the times. But I’ve been using speakers and headphones for most of my 60+ years. And did not own or really see a need for a standalone DAC until just a few years ago.

It’s still kind of an unnecessary luxury imo. But I’ve been happy with the results. I only paid $25 for mine though.

19

The only thing that’s critical with a DAC is that it avoids negative artifacts. My early (cheap) Cirrus Logic and ESS delta sigma DACs whistled and whined. Some higher end ESS DACs have been piercing too. These were extremely easy to identify blind because each had different specific artifacts. My AKM 4490 DAC did away with the bulk of those issues, but not everything.

My entry level ZenDAC with a Burr Brown chip and much more expensive Bifrost 2 avoided piercing whining but have roughness artifacts. Then, the Bifrost 2/64 does not whine and has no audible roughness. I don’t see any point in spending more.

3 Likes
20

For anyone looking to really dive into the myths, @GoldenSound just dropped an excellent video today that is sure to clarify nothing for you.

7 Likes